Reward Prediction Error ([[RPE]]), is thought to be a the activator of learning when ther is a difference between accepted reward and actual reward<ref>[http://www.pnas.org/content/early/2011/03/08/1014269108.full.pdf article 2011 PNAS]</ref>. but it is debated<ref>[http://www.ncbi.nlm.nih.gov/pubmed/17115078?dopt=Abstract article 2007]</ref>.